This is an author version of the contribution published on:
Questa è la versione dell’autore dell’opera:
C.Baggiani, C.Giovannoli, L.Anfossi, C.Passini, P.Baravalle, G.Giraudi: “A connection
between the binding properties of imprinted and non-imprinted polymers: a change of
perspective in molecular imprinting” J.Am.Chem.Soc., 2012, 134(3), 1513-1518 (DOI
10.1021/ja205632t)
The definitive version is available at:
La versione definitiva è disponibile alla URL:
http://pubs.acs.org/journal/jacsat
A connection between the binding properties of imprinted
and non-imprinted polymers: a change of perspective in molecular imprinting
Claudio Baggiani*, Cristina Giovannoli, Laura Anfossi, Cinzia Passini, Patrizia Baravalle†, Gianfranco
Giraudi
Laboratory of Bioanalytical Chemistry, Department of Analytical Chemistry, University of Torino, Torino, Italy.
KEYWORDS. Molecular imprinting, imprinted polymer, non imprinted polymer, molecular recognition, binding isotherm, affinity constant,
binding site density, selectivity
Supporting Information Placeholder
ABSTRACT: In the current paradigm about molecular imprinting, the imprinted binding sites exist as a consequence of the polymerization process around templates, and the properties of non-imprinted polymers (NIPs) has been largely overlooked. Thus, nothing
can be affirmed a priori on the binding properties of NIPs. We propose an alternative view where the imprinting effect is due to the
presence of a template molecule which enhances the pre-existing binding properties of a polymer. If a NIP shows no binding properties towards a target molecule the corresponding imprinted polymer (MIP) will show a weak imprinting effect. On the other hand, if a
NIP shows binding properties towards a target molecule, the corresponding MIP will show a significant imprinting effect. To verify
this hypothesis we prepared a 96-member combinatorial polymeric library in the absence of any template molecule. This library was
screened for several potential ligands and, with no exceptions, the composition of the best binding NIP produced a MIP with excellent binding properties, whereas a low binding NIP formulation produced a MIP with comparable low binding. To validate these results the binding properties towards naproxen and ibuprofen were measured for two combinatorial libraries of polymers, prepared in
the presence (MIP-library) and the absence (NIP-library) of the template molecule. The experiment’s results showed a correlation between the apparent affinity constant measured for the NIP- and the MIP-library, confirming the proposed hypothesis. Moreover, for
closely related molecules, it was shown that binding selectivity is an emergent property derived from the imprinting process, and not a
property of NIPs.
INTRODUCTION
Molecularly imprinted polymers (MIPs) can be obtained by
the polymerization of a mixture of cross-linkers and functional
monomers in the presence of a template dissolved in a proper
porogenic solvent.1 The nature of the resulting material and its
binding properties are influenced not only by the composition
of the pre-polymerization mixture,2 but also by the experimental
conditions employed, such as including the type and the
amount of radical initiator used, the polymerization temperature, the type of polymerization mechanism and so forth.3 It is
often assumed that the template molecule plays a pivotal role,
and that cross-linkers, functional monomers and porogenic
solvents should be chosen by taking into account the chemical
properties of the template. Thus, the current paradigm which
describes the origins of molecular imprinting mechanism can
be illustrated by the well-known empirical model where the
imprinted binding site exists as a direct consequence of the
polymerization of several monomers around the template molecule. This description seems to be confirmed not only by the
huge large amount of papers reported in the last twenty years,
but also by successful in silico simulation of several imprinted
systems.4 Moreover, the existence of imprinted sites is support-
ed by a large amount of experimental data, that indicates how
they act as reversible binding sites with well defined (and surprisingly complex) thermodynamic and kinetic behaviors influenced by steric and electronic features of the template molecule.5
In the current paradigm, there has not been much attention
paid to the properties of non-imprinted polymers (NIP). In fact,
any imprinting effect in a polymer is the consequence of the
presence of the template molecule in the polymerization mixture and its interactions with the mixture components. Thus, it
is very difficult to make reliable predictions about binding
properties of NIPs prepared without any template molecule.
However, this paradigm seems to be challenged in some manner by papers describing MIPs or NIPs with unexpected molecular recognition properties.6 Moreover, several papers have
recently been published about polymers which are characterized
by good selectivity and binding properties towards small organic
molecular targets7 or even larger peptides8 prepared without the
use of a template.
On the basis of these facts, we think that an alternative view
on molecular imprinting is possible. In this hypothesis, illustrated in figure 1, the presence of the template molecule in the
pre-polymerization mixture acts to enhance binding properties
that preexist anyway in a NIP. As a consequence, if a NIP shows
no binding properties towards a target molecule, the corresponding MIP will show a weak imprinting effect, if any. On
the other hand, if the NIP shows binding properties towards a
target molecule, the corresponding MIP will show a significant
imprinting effect.
To verify this hypothesis, in this work we prepared a 96member combinatorial polymeric library in the absence of any
template molecule (NIP-library). This library was screened for
several potential ligands and, with no exception, the composition of the best binding NIP produced a MIP with excellent
binding properties, whereas a low binding NIP formulation
produced a MIP with comparable low binding. To validate
these results, the equilibrium binding properties (affinity constant, binding site density) towards naproxen were measured for
two combinatorial libraries of polymers, prepared in the presence (MIP-library) and the absence (NIP-library) of the template
molecule by varying the functional monomer, the cross-linker
and the porogen. The screening of 96 different polymers confirmed a clear positive correlation between the binding properties measured for the NIP- and the MIP-library.
RESULTS AND DISCUSSION
Synthesis and screening of the polymeric combinatorial library. Our hypothesis of a relationship between the binding
properties of imprinted and non-imprinted polymers was verified by preparing a non-imprinted library of 96 elements and
screening it for the binding of several ligands. After that, the
best binding non-imprinted polymers were compared with the
related imprinted polymers.
To assure a large degree of molecular diversity in the composition of the polymers, we combined very different functional
monomers, cross-linkers and porogenic solvents, all previously
reported in literature as components of successful molecularly
imprinted
polymers.9
Neutral
(acrylamide,
2hydroxyethylmethacrylate), acid (methacrylic acid) and basic (4vinylpyridine) compounds were used as functional monomers,
while cross-linkers were selected in terms of the number of possible polymerizable groups: two (divinylbenzene, ethylene dimethacrylate, glycerol dimethacrylate), three (pentaerithrytole
triacrylate, trimethylolpropane trimethacrylate) and four (pentaerithrytole tetraacrylate). Porogenic solvents were selected in a
way so as to represent different typologies of organic solvent:
with aromatic (toluene), hydrophobic (chloroform) and hydrophilic (acetonitrile and tetrahydrofurane) character.
In an attempt to assure that any relationship between the
molecular recognition properties of imprinted and nonimprinted polymers was not the spurious effect of chance, the
non-imprinted library was screened in such a way to be sure
that the degree of molecular diversity in the ligand structures
was sufficiently wide. Chloramphenicol and cortisol are neutral
molecules, while diclofenac, ibuprofen and naproxen are acids
with pK of about 4-5, bisphenol A and theophylline very weak
acids, and metribuzin and pyrimethanil weak basis. The hydrophobicity covers a very large interval of logP values, ranging
from -0.02 for theophylline – an essentially hydrophilic molecule – to 4.51 for diclofenac, that in fully protonated form is
very hydrophobic. Moreover, all the considered ligands have
been previously reported in literature as template molecules,
and the corresponding imprinted polymers have been largely
studied and the binding behavior is very well known.10
The effect of the large molecular diversity represented by the
panel of ligands is well illustrated in figure 2, where the box
plot reports the spreading of the bound-to-free (B/F) ratio values measured for each of the ligands. It is possible to see that
different ligands bind in very different ways, with B/F values
between 0.05 and 0.5 (1st-3rd percentile), ranging from results
dispersed at wide intervals of B/F values (diclofenac and pyrimethanil) to results present at relatively narrow intervals of
B/F values (chloramphenicol, cortisol, metribuzin and theophylline), with some intermediate situations (bisphenol A, ibuprofen, naproxen and pyrimethanil).
Comparison of the binding properties of imprinted and
non-imprinted polymers. The B/F results related to the different ligands were examined to identify the composition of the
best and worst binding polymers for each of the ligands considered and the ligand binding has been measured for the corresponding imprinted polymers. In table 1 the B/F ratio for each
of the polymer pairs are reported. Despite the difficulty to exactly compare binding data when the B/F ratio assumes extreme values (B/F<0.1 or B/F>10), from these results it is nevertheless possible to observe that, with no exception, the composition of the best binding NIP produced a MIP with excellent
binding properties, characterized by a marked increase of the
ligand binding (evalued as the increase of the difference between B/F measured for NIP and MIP), whereas a low binding
NIP formulation produced a MIP with comparable low ligand
binding. Interestingly, it seems that the pair 4-vinylpyridine /
divinylbenzene represent the optimal functional monomer /
cross-linker combination, as it is present in 5 of 9 formulations
corresponding to high binding polymers (polymers binding
bisphenol A, cortisol, diclofenac, ibuprofen and naproxen), and
that 4-vinylpyridine – but no divinylbenznene - is present in
other 2 formulations (polymers binding chloramphenicol and
metribuzin). This result can be related to the fact that 6 out of 9
tested templates are molecules with carboxyl or hydroxyl substituents, known to interact with the pyridine ring through
hydrogen bond or ion pair interactions,10 and that metribuzin,
a weak acidic molecule, is both a good hydrogen bond acceptor
and donor, able to interact with 4-vinylpyridine – a strong hydrogen bond acceptor. On the other hand, it seems impossible
to clearly identify a functional monomer / cross-linker combination typical of formulations giving poorly binding polymers.
Comparison of the binding isotherms of imprinted and
non-imprinted polymers. The measurement of the B/F ratio
for MIPs and NIPs reported in the previous section is related to
a single point in a binding isotherm, measured for a ligand
concentration of 50 g/ml. Thus, only indirect information on
the binding properties of the polymers can be obtained. To
better validate these results, it was decided to gather direct information on the ligand binding properties (i.e. apparent affin-
ity constant, Keq, and binding site density, Bmax) by measuring
the whole binding isotherm for two combinatorial libraries of
polymers, prepared in the presence (MIP-library) and the absence (NIP-library) of naproxen. Despite the well-known complexity of the binding behavior of MIPs,11 a simple Langmuir
model was chosen to limit the number of the experimental
points necessary to obtain accurate estimates of equation parameters.
The comparison of Keq values measured on the MIP- and
NIP-libraries applying a Mann-Whitney rank sum test (figure 3)
shows that the numerical difference between the two groups is
greater than would be expected by chance (P<0.000001), thus
confirming that there is a statistically significant difference between the distribution of Keq values in the MIP- and NIPlibraries. From the plot reported in figure 4, it is possible to
observe a statistically significant direct relationship between the
Keq values of MIPs and NIPs, expressed by a linear regression
model of Keq(MIP) vs. Keq(NIP):
Keq(MIP)naproxen = 0.298(±0.753) + 1.39(±0.0832)Keq(NIP)naproxen
r2=0.748, n=96, s=3.29, F=278.4, P<0.0001
(2)
It should be noted that the slope of the regression line is
greater than the unit, indicating that not only there is a marked
difference between Keq values measured for the NIP- and MIPlibrary, but that Keq values measured for the MIP-library increases proportionally with the increase of the Keq values for the
NIP-library.
As regards Bmax, the comparison of the values measured on
the MIP- and NIP-libraries (figure 5) with the same test used for
Keq shows that the numerical difference between the two groups
is not greater than would be expected by chance (P=0.1426),
confirming that there is not a significant difference between
Bmax measured for the NIP- and MIP-library. Thus, it seems that
the main difference between MIPs and NIPs will be related to
differences in the magnitude of binding affinity, and not to the
number of available binding sites.
As in the case for Keq(MIP) vs. Keq(NIP) model, a statistically
significant linear regression of Bmax(MIP) vs. Bmax(NIP), whose
plot is reported in figure 6, is described in the following equation:
Bmax(MIP) naproxen = 4.53(±3.90) + 1.10(±0.0624)Bmax(NIP) naproxen
r2=0.768, n=96, s=21.9, F=311.6, P<0.0001.
(3)
Considering that plots reported in figure 4 and 6 show the
presence of linear models correlating the binding properties of
NIPs and MIPs, it is clear that there are many low- Keq, low-Bmax
MIPs corresponding to low-Keq, low-Bmax NIPs and a more limited number of high-Keq, high-Bmax MIPs corresponding to highKeq, high-Bmax NIPs, but there are no high-Keq, high-Bmax MIPs
with compositions corresponding to low-Keq, low-Bmax. This
confirms our working hypothesis, i.e. that if a NIP shows limited
binding properties towards a target molecule, the corresponding MIP
will show a weak imprinting effect, if any. On the contrary, if the NIP
shows marked binding properties towards a target molecule, the corresponding MIP will show a significant imprinting effect.
Binding selectivity of imprinted and non-imprinted polymers. Naproxen was chosen as an imprint molecule as it was
possible to compare its binding properties to ibuprofen, a closely related ligand already examined in the preliminary screening
of the NIP library. Thus, binding selectivity was studied by
comparing the measured values of Keq, and Bmax of naproxen
and ibuprofen on NIP- and (naproxen-imprinted)MIP-libraries.
The statistical comparison of Keq values measured for ibuprofen on the (naproxen-imprinted)MIP- and NIP-libraries (figure 7) shows that the numerical difference between the two
groups of data is greater than would be expected by chance
(P=0.016), thus confirming that also for ibuprofen there is a
statistically significant difference between the distribution of Keq
values in the (naproxen-imprinted)MIP- and NIP-libraries and
that this difference can be attributed to the recognition of the
ibuprofen molecules by the imprinted library. On the contrary,
the comparison of the Bmax values measured for ibuprofen on
the (naproxen-imprinted)MIP- and NIP-libraries with the same
test used for Keq values (figure 8) shows that the numerical difference between the two groups is not greater than would be
expected by chance (P=0.284), confirming what was observed
for naproxen: there is not a significant difference between Bmax
values measured for the NIP- and MIP-library.
As polymer selectivity seems to be controlled by the ligand affinity only, while the number of binding sites seems to be unimportant, it is interesting to make a direct comparison of the
corresponding linear regression models of Keq(ibuprofen) vs.
Keq(naproxen) calculated for NIP- and MIP-libraries:
Keq(NIP)naproxen = 1.24(±0.481) + 0.926(±0.058)Keq(NIP)ibuprofen
r2=0.731, n=96, s=2.12, F=255.4, P<0.0001
(4)
Keq(MIP)naproxen = 0.0114(±0.562) + 1.28(±0.0735)Keq(MIP)ibuprofen
r2=0.764, n=96, s=3.19, F=303.9, P<0.0001
(5)
From the plot reported in figure 9, it is possible to observe
that the numerical value for the slope of the regression model
calculated for the NIP-library (eq.4) is about one, indicating that
naproxen and ibuprofen show the same binding behavior and
are recognized in the same manner by the NIP-library. On the
contrary, the regression model calculated for the (naproxenimprinted)MIP-library (eq.5) shows a slope significantly greater
than one, indicating that naproxen is better recognized than
ibuprofen. Thus, by analogy to what is known about the capabilities of racemic resolution typical of MIPs imprinted against
optically active molecules, it can be assumed that, for closely
related molecules, the binding selectivity seems to be a molecular recognition property arising from the imprinting process.
related molecules is an emergent property derived from the
imprinting process, and NIPs tend to be poorly selective.
Apart from the contribution to a better understanding of the
fundamentals of molecular imprinting, we think that these
results have some important practical implications in MIP
technology. In fact, as NIP- and MIP-libraries show the same
binding behavior, it should be possible with a reasonable rate of
success (thus not excluding that some false positives and false
negatives happens) to identify efficient prepolymerization mixtures to prepare high binding imprinted polymers simply by
screening a NIP-library, thus making easier the cumbersome
process of optimizing a MIP formulation. In fact, not only is the
synthesis of a NIP-library much cheaper and simpler than a
MIP-library, as no template has to be used to imprint the polymers and subsequently extracted, but the same library can be
recycled many times, to screen for different target ligands, simultaneously or in sequence, without the need to prepare many
different MIP-libraries. Moreover, the relatively simple accessibility to very large libraries of hundreds of different polymers
paves the way to fast screening for exotic polymer formulations
involving functional monomers and cross-linker which are
much more different than the “classical” methacrylic acid, 4vinylpyridine or ethylene dimethacrylate.
CONCLUSIONS
The libraries considered in this work can be considered representative of widely used experimental conditions involving
small molecules as templates and non-covalent bulk imprinting
conditions. Thus, as the current study is not concerned with
other different imprinting approaches (like covalent imprinting,
ion imprinting, use of large templates like proteins and so on),
we think that our results can be considered valid and of general
value for the non-covalent imprinting approach. The clear and
positive correlation between the apparent affinity constants
measured both for the NIP- and the MIP-libraries means that
these libraries share the same binding behavior, confirming our
initial hypothesis; in the imprinting process the presence of the
template molecule in the pre-polymerization mixture acts to
enhance the resulting MIP binding properties that exist in the
corresponding NIP.
As regard the selectivity of the molecular recognition properties, considering strictly related ligands – as in the case for the
pair naproxen / ibuprofen - the experimental results reported
here confirm what common knowledge of the imprinting process is: selectivity between enantiomeric pairs or structurally
TABLES.
Table 1: bound to free ratio measured for selected polymers presenting the best and the worst ligand binding in according with
the binding screening of the non imprinted polymeric library
Best binding polymer
Ligand
Polymer formulation
Worst binding polymer
B/F
B/F
MIP
NIP
Polymer formulation
B/F
B/F
MIP
NIP
Bisphenol A
4VP-DVB-CHCl3
1.49
0.95
HEMA-DVB-MeCN
0.02
0.02
Chloramphenicol
4-VP-PETA- CHCl3
0.63
0.43
MAA-PETA-TOL
0.02
0.02
Cortisol
4VP-DVB- CHCl3
1.37
0.55
HEMA-PETA-TOL
0.07
0.05
Diclofenac
4VP-DVB-MeCN
32.1
7.84
MAA-PETA-TOL
0.07
0.06
Ibuprofen
4VP-DVB-THF
1.69
1.06
AM-GDMA-TOL
0.03
0.02
Metribuzin
4VP-GDMA-THF
0.61
0.42
MAA-DVB-MeCN
0.13
0.11
Naproxen
4VP-DVB-TOL
1.69
0.91
HEMA-EDMA-TOL
0.08
0.05
Pyrimethanil
MAA-DVB-THF
22.8
4.84
HEMA-GDMA-TOL
0.02
0.01
Theophylline
AM-DVB- CHCl3
0.53
0.37
MAA-TRIM-MeCN
0.04
0.02
FIGURES.
Figure 1. The working hypothesis: the presence of the template
molecule in the pre-polymerization mixture acts to enhance binding properties that anyway exist in a NIP. Thus, if a NIP shows
limited binding properties towards a target molecule, the corresponding MIP will show a weak imprinting effect, if any. On the
contrary, if the NIP shows marked binding properties towards a
target molecule, the corresponding MIP will show a significant
imprinting effect.
Figure 3. Comparison of apparent affinity constant (Keq) values
measured for naproxen on the MIP- and NIP-libraries applying a
Mann-Whitney rank sum test. See supporting informations (note
S1) for the statistical meaning of this plot.
Figure 2. Bound-to-free ratio (B/F) values measured for each of the
ligands by overnight incubation at 4 °C of 10 mg of polymer suspended in 200 l of 50 g/ml ligand solution in acetonitrile. See
supporting informations (note S1) for the statistical meaning of
this plot.
Figure 4. Relationship between the apparent affinity constant (Keq)
values measured for naproxen on the MIP- and NIP-libraries. The
red line indicates the linear regression model of Keq(MIP) vs.
Keq(NIP). The black line represents the upper edge for the Keq(MIP)
< Keq(NIP) region.
Figure 5. Comparison of the binding site density (Bmax) values
measured for naproxen on the MIP- and NIP-libraries applying a
Mann-Whitney rank sum test.
Figure 7. Comparison of apparent affinity constant (Keq) values
measured for ibuprofen on the (naproxen-imprinted)MIP- and NIPlibraries applying a Mann-Whitney rank sum test.
Figure 6. Relationship between the binding site density (Bmax) values for naproxen measured on the MIP- and NIP-libraries. The blue
line indicates the linear regression model of Bmax(MIP) vs.
Bmax(NIP). The black line represents the upper edge for the
Bmax(MIP) < Bmax(NIP) region.
Figure 8. Comparison of the binding site density (Bmax) values
measured for ibuprofen on the (naproxen-imprinted)MIP- and NIPlibraries applying a Mann-Whitney rank sum test.
Supporting Information. Materials and methods, template structures, composition of the polymeric combinatorial libraries. This
material is available free of charge via the Internet at
http://pubs.acs.org.”
AUTHOR INFORMATION
Corresponding Author
* Claudio Baggiani, Department of Analytical Chemistry – University of Torino, Via Pietro Giuria 5, 10125 Torino – Italy. email:
[email protected]
Present Addresses
† Chemical Control s.r.l, Via Celdit 2, 12100 Madonna dell’Olmo
– Cuneo, Italy.
AKNOWLEDGMENT
C.B. would like to thank Prof. Gunther Wulff for his very helpful
advice.
Figure 9. Relationships between the apparent affinity constant
(Keq) values measured for ibuprofen and naproxen on the MIP- (red
circles) and NIP-libraries (blue circles). The continuous lines indicate the linear regression models for Keq(ibuprofen) vs.
Keq(naproxen) calculated for NIP- (blue line) and MIP-libraries (red
lines).
REFERENCES
(1)
(2)
(3)
(4)
(5)
ASSOCIATED CONTENT
(a) Molecularly Imprinted Polymer-Man made Mimics of Antibodies
and Their Applications in Analytical Chemistry; Sellergren, B., Ed.;
Elsevier: Amsterdam, 2001. (b) Molecular Imprinting: From Fundamentals to Applications; Komiyama, M., Ed.; Wiley-VCH:
Weinheim, 2002. (c) Molecularly Imprinted Materials: Science and
Technology; Yan, M., Ramström, O., Eds.; Dekker: New York,
2005.
(a) Ansell, R. J.; Kuah, K. L. Analyst 2005, 130, 179-187. (b)
Ansell, R. J.; Wang, D. Y.; Kuah, J. K. L. Analyst 2008, 133,
1673-1683. (c) Ansell, R. J.; Wang, D. Y. Analyst 2009, 134,
564-576. (d) Zhang, Y. G.; Song, D.; Lanni, L.M.; Shimizu,
K.D. Macromolecules 2010, 43, 6284-6294.
(a) Piletsky, S. A.; Piletska, E. V.; Karim, K.; Freebairn, K. W.;
Legge, C. H.; Turner, A. P. F. Macromolecules 2002, 35, 74997504. (b) Navarro-Villoslada, F.; San Vicente, B.; MorenoBondi, M. C. Anal. Chim. Acta 2004, 504, 149-162. (c) Piletsky,
S. A.; Guerreiro, A.; Piletska, E. V.; Chianella, I.; Karim, K.;
Turner, A. P. F. Macromolecules 2004, 37, 5018-5022. (d) Lu,
Y.; Li, C. X.; Wang, X. D.; Sun, P. C.; Xing, X. H. J. Chromatogr. B 2004, 804, 53–59. (e) Piletsky, S. A.; Mijangos, I.; Guerreiro, A. R.; Piletska, E. V.; Chianella, I.; Karim, K.; Turner, A.
P. F. Macromolecules 2005, 38, 1410-1414. (f) Piletska, E. V.;
Guerreiro, A. R.; Whitcombe, M. J.; Piletsky, S. A. Macromolecules 2009, 42, 4921–4928.
Nicholls, I. A.; Andersson, H. S.; Charlton, C.; Henschel, H.;
Karlsson, B. C. G.; Karlsson, J. G.; O’Mahony, J.; Rosengren,
A. M.; Rosengren, K. J.; Wikman, S. Biosens. Bioelectron. 2009,
25, 543–552.
(a) Chen, Y.; Kele, M.; Quinones, I.; Sellergren, B.; Guiochon,
G. J. Chromatogr. A, 2001, 927, 1–17. (b) Umpleby, R. J.; Baxter, S. C.; Rampey, A. M.; Rushton, G. T.; Chen, Y. Z.; Shimizu, K. D. J. Chromatogr. B 2004, 804, 141-149. (c) Baggiani, C.;
Giraudi, G.; Giovannoli, C.; Tozzi, C.; Anfossi, L. Anal. Chim.
Acta 2004, 504, 43–52. (d) Kim, H.; Guiochon, G. J. Chromatogr. A 2005, 1097, 84–97. (e) Kim, H.; Kaczmarski, K.; Guiochon, G. J. Chromatogr. A 2006, 1101, 136–152. (f) Garcia-
(6)
(7)
(8)
Calzon, J. A.; Diaz-Garcia M. E. Sens. Actuat. B 2007, 123,
1180-1194. (g) Lee, W. C.; Cheng, C. H.; Pan, H. H.; Chung,
T. H.; Hwang, C. C. Anal. Bioanal. Chem. 2008, 390, 11011109.
(a) Martin, P. D.; Wilson, T. D.; Wilson, I. D.; Jones, G. R.
Analyst 2001, 126, 757–759. (b) Zhou, S. N.; Lai, E. P. C. React. Funct. Polym. 2004, 58, 35–42. (c) Anfossi, L.; Baggiani, C.;
Baravalle, P.; Giovannoli, C.; Guzzella, L.; Pozzoni, F. Anal.
Lett. 2009, 342, 807–820.
(a) Breton, F.; Rouillon, R.; Piletska, E. V.; Karim, K.; Guerreiro, A.; Chianella, I.; Piletsky, S. A. Biosens. Bioelectron. 2007,
22, 1948–1954. (b) Pascale, M.; De Girolamo, A.; Visconti, A.;
Magan, N.; Chianella, I.; Piletska, E. V.; Piletsky, S. A. Anal.
Chim. Acta 2008, 609, 131–138. (c) Guerreiro, A.; Soares, A.;
Piletska, E.; Mattiasson, B.; Piletsky, S. Anal. Chim. Acta 2008,
612, 99-104. (d) Piletska, E.; Karim, K.; Coker, R.; Piletsky, S.
J. Chromatogr. A 2010, 1217, 2543-2547. (e) Piletska, E.V.;
Stavroulakis, G.; Karim, K.; Whitcombe, M. J.; Chianella, I.;
Sharma, A.; Eboigbodin, K. E.; Robinson, G. K.; Piletsky, S. A.
Biomacromolecules 2010, 11, 975–980.
Hoshino, Y.; Haberarcker, W.W.; Kodama, T.; Zeng, Z.;
Okahata, Y.; Shea, K.J. J. Am. Chem. Soc. 2010, 132, 1364813650.
SYNOPSIS TOC
(9)
(10)
(11)
(a) Cormack, P. A. G.; Elorza, A. Z. J. Chromatogr. B, 2004,
804, 173–182. (b) Mayes, A. G.; Whitcombe, M.,J. Adv. Drug
Deliv. Rev. 2005, 57, 1742–1778.
Bisphenol A: (a) Navarro-Villoslada, F.; San Vicente, B.; Moreno-Bondi, M. C. Anal. Chim. Acta 2004, 504, 149-162. Chloramphenicol: (a) Levi, R.; McNiven, S.; Piletsky, S. A.;
Cheong, S. H.; Yano, K.; Karube, I. Anal. Chem. 1997, 69,
2017-2021. (b) Baggiani, C.; Baravalle, P.; Giovannoli, C.; Anfossi, L.; Giraudi, G. Biosens. Bioelectron. 2010, 26, 590-595.
Diclofenac: (a) Fernandez-Llano, L.; Blanco-Lopez, M. C.; Lobo-Castanon, M. J.; Miranda-Ordieres, A. J.; Tunon-Blanco, P.
Electroanalysis 2007, 19, 1555-1561. Ibuprofen: (a) Farrington,
K.; Regan, F. Biosens. Bioelectron. 2007, 22, 1138-1146.
Metribuzin: (a) Zhang, S. J.; Yang, G. L.; Zheng, Z. S.; Chen, Y.
Chromatographia 2009, 69, 615-619. Naproxen: (a) O'Mahony,
J.; Karlsson, B. C. G.; Mizaikoff, B.; Nicholls, I. A. Analyst
2007, 132, 1161-1168. Pyrimethanil: (a) Baggiani, C.; Baravalle, P.; Anfossi, L.; Tozzi, C. Anal. Chim. Acta 2005, 542, 125134. Theophylline: (a) Vlatakis, G.; Andersson, L. I.; Muller,
R.; Mosbach, K. Nature 1993, 361, 645-647.
Garcia-Calzon, J.A.; Diaz-Garcia, M. E. Sens. Actuat. B, 2007,
123, 1180–1194.